Dynamic execution of predictive models and workflows
Disclosed herein are systems, devices, and methods related to assets and predictive models and corresponding workflows that are related to the operation of assets. In particular, examples involve defining and deploying aggregate, predictive models and corresponding workflows, defining and deploying individualized, predictive models and/or corresponding workflows, and dynamically adjusting the execution of model-workflow pairs.
1. A local analytics device that is capable of being physically coupled to an asset, the local analytics device comprising:
an asset interface configured to communicatively couple the local analytics device to one or more on-board components of the asset;
a network interface configured to facilitate communication with a remote computing system over a communication network;
at least one processor;
a non-transitory computer-readable medium; and
program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor to cause the local analytics device to perform functions comprising:
locally executing a predictive model that is related to the operation of the asset, wherein the predictive model was defined based on operating data for a plurality of assets;
while locally executing a predictive model, detecting an indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system;
in response to detecting the indication, reducing the responsibility of the local analytics device with respect to execution of the predictive model; and
operating in accordance with the reduced responsibility.
2. The local analytics device of claim 1 , wherein locally executing the predictive model for the asset comprises:
applying the predictive model to operating data that is received from the asset via the asset interface.
3. The local analytics device of claim 1 , wherein detecting the indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system comprises:
detecting a reduction in processing capabilities of the local analytics device.
4. The local analytics device of claim 1 , wherein detecting the indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system comprises:
detecting a change in a characteristic of the communication network over which the local analytics device communicates with the remote computing system.
5. The local analytics device of claim 4 , wherein the characteristic of the communication network is a signal strength.
6. The local analytics device of claim 1 , wherein detecting an indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system comprises:
receiving, from the remote computing system, a message indicating that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system.
7. The local analytics device of claim 1 , wherein reducing the responsibility of the local analytics device with respect to execution of the predictive model comprises:
ceasing any local execution of the predictive model for the asset.
8. The local analytics device of claim 1 , further comprising program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor to cause the local analytics device to perform functions comprising:
in response to detecting the indication, sending the remote computing system an instruction to take over responsibility for executing the predictive model for the asset.
9. The local analytics device of claim 1 , wherein the predictive model being locally executed by the local analytics device comprises one portion of a distributed predictive model.
10. The local analytics device of claim 1 , wherein the predictive model related to the operation of the asset comprises a predictive model that is configured to (a) receive operating data for the asset and (b) output an indication of whether a failure of at least one given type is likely to occur at the asset within the foreseeable future.
11. The local analytics device of claim 1 , further comprising program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor to cause the local analytics device to perform functions comprising:
before locally executing the predictive model, receiving the predictive model over a communication network from the remote computing system.
12. A non-transitory computer-readable medium, wherein the non-transitory computer-readable medium has been provisioned with software that is executable to cause a local analytics device that is (a) physically coupled to an asset and (b) communicatively coupled to one or more on-board components of the asset via an asset interface of the local analytics device to perform functions comprising:
locally executing a predictive model that is related to the operation of the asset, wherein the predictive model was defined based on operating data for a plurality of assets;
while locally executing the predictive model, detecting an indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to a remote computing system that is communicatively coupled to the local analytics device over a communication network;
in response to detecting the indication, reducing the responsibility of the local analytics device with respect to execution of the predictive model; and
operating in accordance with the reduced responsibility.
13. The non-transitory computer-readable medium of claim 12 , wherein beginning to locally execute the predictive model for the asset comprises:
applying the predictive model to operating data that is received from the asset via the asset interface.
14. The non-transitory computer-readable medium of claim 12 , wherein detecting the indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system comprises one of (a) detecting a reduction in processing capabilities of the local analytics device or (b) detecting a change in a characteristic of a communication network that communicatively couples the local analytics device and the remote computing system.
15. The non-transitory computer-readable medium of claim 12 , wherein detecting an indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system comprises:
receiving, from the remote computing system, a message indicating that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system.
16. The non-transitory computer-readable medium of claim 12 , wherein reducing the responsibility of the local analytics device with respect to execution of the predictive model comprises:
ceasing any local execution of the predictive model for the asset.
17. The non-transitory computer-readable medium of claim 12 , wherein the software is further executable to cause the local analytics device to perform the following function:
in response to detecting the indication, sending the remote computing system an instruction to take over responsibility for executing the predictive model for the asset.
18. A computer-implemented method, the method comprising:
locally executing, at a local analytics device that is (a) physically coupled to an asset and (b) communicatively coupled to one or more on-board components of the asset via an asset interface of the local analytics device, a predictive model that is related to the operation of the asset, wherein the predictive model was defined based on operating data for a plurality of assets;
detecting an indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to a remote computing system that is communicatively coupled to the local analytics device over a communication network;
in response to detecting the indication, reducing the responsibility of the local analytics device with respect to execution of the predictive model; and
operating the local analytics device in accordance with the reduced responsibility.
19. The computer-implemented method of claim 18 , wherein detecting the indication that at least some responsibility for executing the predictive model is to be shifted from the local analytics device to the remote computing system comprises one of (a) detecting a reduction in processing capabilities of the local analytics device or (b) detecting a change in a characteristic of the communication network that communicatively couples the local analytics device and the remote computing system.
20. The computer-implemented method of claim 18 , wherein reducing the responsibility of the local analytics device with respect to execution of the predictive model comprises:
ceasing any local execution of the predictive model for the asset.